Have feedback? Can't find your answer in our Help pages?
eBook File Formatting Tips
Microsoft WordThe initial formatting you chose in Word may be slightly altered to conform to Kindle device specifications. Microsoft Word format comes with a few quirks and not all file format details are well documented.
- Lists in Microsoft Word documents, especially nested lists with subcategories, don't convert well.
- If you change list numbering between numerals and letters (e.g., I, II, III, a, b, c, 2.1.3), the conversion may also provide less than satisfying results.
- If you use tables of contents in your files, page numbers will not be updated correctly. This is because the text size and page numbers in a reflowable book don't mean the same thing as in a fixed-layout book. Make sure to use the standard Microsoft Word heading styles and Table of Contents insertion for best results.
- If you're using the Microsoft Word Equation tool, not all mathematical notations will convert well. We suggest converting the equation into an image (recommended: 300 DPI PNG).
- Don't upload your content with Track Changes on.
HTMLAvoid frames. Some electronic books have a frame (e.g., on the left side of the page) with bookmarks or chapter information, and the content in a frame on the right. If you upload HTML, the ideal content is in a single HTML file, with references to additional image files in the same folder (zipped up).
You can use the <table> tag to create tables. If you have a large table, split it into multiple tables for best results.
For more information on formatting your tables, see section 3.7 of the Kindle Publishing Guidelines.
Kindle Create allows you to add interactive content to eBooks, such as audio, video, images, and hyperlinks.
TXTFormat all text into paragraphs. Any hard line breaks that you insert into a text file (pressing "Enter") will make your text display poorly. Use two line breaks (press "Enter" twice) to separate lines of paragraphs.
- Any HTML tags inside a text file will be ignored and displayed as plain text, showing the actual HTML code. You can't mix HTML and plain text.
- Even though plain text using unusual characters and glyphs may display fine on your computer, it won't necessarily convert well to Kindle. Check the list of supported characters to make sure.
You can also try using Kindle Create. This is a downloadable tool that helps you convert PDFs and allows you to add interactive content, such as audio, video, images, and hyperlinks. Learn more about Kindle Create.
If you're unable to save the content as HTML, try to export or save the content in a format where you can open it with Microsoft Word, and save it as HTML from there. If you can't get your content out of the PDF format, refer to these troubleshooting tips:
- The exact layout of my PDF was lost. Conversion usually can't retain exact formatting because PDF is a fixed layout format that is designed for printing. Amazon KDP uses a reflowable format that is designed to be displayed on screens that differ greatly in size and width/height ratio. For an explanation on reflowable versus fixed layout, see eBook Conversion Formats.
- My formatting was lost. The PDF format is a "destructive" format in the sense that you lose all high-level information. PDFs store characters and coordinates for each character or word. The information contained is very much like a scan of the printed pages. The notions of words, lines, and paragraphs don't exist in the PDF format, much less the flow of text (what follows what). To convert the PDF file into a reflowable format that is viewable on different screen sizes using any font size, our software has to apply "Artificial Intelligence" like algorithms that simulate the way a human reader with knowledge of western typographical conventions would read the text. These algorithms reassemble characters into words, words into lines, lines into paragraphs, and paragraphs into columns. Even though we use a lot of complex algorithms, some of the formatting is still lost. The goal of our software is to extract the text from the PDF without corrupting it and retain a minimum of formatting, which is already an achievement.
- My tables didn’t convert. As mentioned in the previous paragraph, information about paragraphs, columns, and tables is lost in PDF files. For instance, tables consist of lines drawn at various locations on the page. As of today, our software is unable to determine that the lines drawn on the page form a table.
- Images are not the same size as in the PDF. The pictures are not the same size either because they had different DPIs (dots per inch, or screen resolution) in the PDF, or because the conversion process had to resize them to keep the size of the file reasonable.